Heterogeneous nonuse values for Arches and Zion National Parks

This paper identifies several kinds of heterogeneity in nonuse values using a contingent choice experiment about two national parks in Utah. Such heterogeneity has potentially important implications for national park management. When management resources are scarce, it matters that nonuse values differ across park resources or attributes. Resources that generate more nonuse values deserve more protection, including the impacts of visitor activities. Heterogeneity by location implies that management decisions appropriate for one park can’t be adopted in other parks without careful consideration. Heterogeneity of nonuse values matters for management if there are political, equity, or justice reasons to pay more attention to particular groups in society. Heterogeneity by type of nonuse value matters for management, too: existence values have different implications than bequest values since the latter still are generated by (future) visitation. The contingent choice experiment used in this paper asks respondents to compare different potential future scenarios for management of either Zion or Arches National Park. Each scenario is made up of potential changes to seven attributes: wilderness protection, educational outreach, animal conservation, plant conservation, visitation levels, cultural/historical protection, and an annual tax payment. Respondents’ rankings are used to estimate a variety of mixed logit models. In addition, respondents are asked to report what percentage of the total value they receive from each attribute is due to use values, option values, bequest values, and existence values. While some previous researchers used similar questions to decompose nonuse values, this paper uses the decomposition questions differently. First, they are asked separately about each attribute in the experimental design of a contingent choice survey, as opposed to being asked once about total value or total nonuse value. Second, they are used as part of a mixed logit estimation of the relative part-worths of each attribute, rather than being used to decompose the final valuation estimates. Results indicate substantial heterogeneity by park attribute, location, individual, and type of nonuse value. These differences across attributes are largely independent of demographic and socioeconomic variables, though. The paper concludes with a discussion of the implications of the empirical findings for park management.